Quote
We'll have National Holiday from Oct. 1 to 7, shipping will resume after we're back to work.
If my miners aren't shipped by then I am going to be so pissed!
ufo
Topic: ANTMINER S7 is available at bitmaintech.com with 4.86TH/s, 0.25J/GH - page 345. (Read 527816 times)
September 29, 2015, 09:59:08 AM
If my miners aren't shipped by then I am going to be so pissed!
ufo
September 29, 2015, 09:55:27 AM
They are not the same.
Efficiency will remain the same as shown, however Power consumption is reduced in the 4.66 TH/s unit.
Rich
Quote
1. Hash Rate: 4.86 TH/s ±5%
2. Power Consumption: 1210 W (at the wall, with APW3, 93% efficiency, 25C ambient temp)
3. Power Efficiency: 0.25 J/GH (at the wall, with APW3, 93% efficiency, 25°C ambient temp)
2. Power Consumption: 1210 W (at the wall, with APW3, 93% efficiency, 25C ambient temp)
3. Power Efficiency: 0.25 J/GH (at the wall, with APW3, 93% efficiency, 25°C ambient temp)
Quote
1. Hash Rate: 4.66 TH/s ±5%
2. Power Consumption: 1160 W (at the wall, with APW3, 93% efficiency, 25C ambient temp)
3. Power Efficiency: 0.25 J/GH (at the wall, with APW3, 93% efficiency, 25°C ambient temp)
2. Power Consumption: 1160 W (at the wall, with APW3, 93% efficiency, 25C ambient temp)
3. Power Efficiency: 0.25 J/GH (at the wall, with APW3, 93% efficiency, 25°C ambient temp)
Efficiency will remain the same as shown, however Power consumption is reduced in the 4.66 TH/s unit.
Rich
September 29, 2015, 09:51:02 AM
I suspect they have some units with an unacceptable HW error rate at 600MHz so have had to drop the Frequency to 575MHz. Very reminiscent of the S3.
Rich
Rich
I'm not quite sure, i noticed that beside the header text, all the rest of the info is the same, same J/GH, but also same power draw, so either its just a different clock and they're selling them higher or the Batch 3 just run the higher freq better while also taking the same J/GH.
One of those value is thus wrong. Either the power draw has to be different or the J/GH.
September 29, 2015, 09:48:39 AM
Bitmain has updated their site. The AntRouter test is gone, and they've added batch 2 and 3 of the S7.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
I also noticed than and am going through the pages to try to find an explanation...
Maybe it will just come clocked higher? Or its a "v1.1" that has more efficient hardware/software support?
Rich
September 29, 2015, 09:47:03 AM
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
There might be a slight issue with that:
Quote
We'll have National Holiday from Oct. 1 to 7, shipping will resume after we're back to work.
(Bitmain home page).
So you might be getting a Batch 2 miner at a Batch 1 price.
September 29, 2015, 09:44:46 AM
Bitmain has updated their site. The AntRouter test is gone, and they've added batch 2 and 3 of the S7.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
I also noticed that and am going through the pages to try to find an explanation...
Maybe it will just come clocked higher? Or its a "v1.1" that has more efficient hardware/software support?
September 29, 2015, 09:31:27 AM
Bitmain has updated their site. The AntRouter test is gone, and they've added batch 2 and 3 of the S7.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
Interesting that batch 2 is 4.66TH/s for $1585 and batch 2 is 4.86TH/s for $1658. Both ship in the same date range.
So I guess you pay a little more for better performance? Have they ever done that before?
In the past separate batches would have separate ship date ranges, not slightly different specs.
I just wish they'd ship the rest of batch 1. Still waiting for my second order from Sept 5.
September 29, 2015, 09:29:09 AM
I see the difference on controller on PCI-e going up.
Is there more details as far as the housing? I guess I would love to see side by sides if anyone happens to have both.
2 picture says more than 100 dogs bark
Click to enlarge
Click to enlarge
"New" S5+ and "New" S7
September 29, 2015, 08:49:43 AM
Can somebody tell me what the effect is of a high HW? If you have two different miners at the same speed, they hash about the same but one has a much higher HW. And on the pool there is also not much effect.
This beg the question of what "high Hardware error" means to you. You can often save something like 5-10% in power just by fine tuning the voltage controls. So unless you're talking about 7% Hardware error like i have on my Zeus miners, you're probably fine. I have some Ants with 0.0001% Error, some other at the same settings do 0.05% error. I even have some S1'S that go up to 2% HWE.
I don't think you should see 0.1%+ on the S7 even if you push it however.
September 29, 2015, 08:44:03 AM
Wonder when bitmain will start shipping S7s again. My Sept. 2nd order still hasn't changed to shipped, and I just noticed that they removed the S7 from the sales page.
September 29, 2015, 08:40:09 AM
Up to a point HW errors are not a problem. In fact having 0% HW errors although very nice means that in practice you would be better running the miner at a lower voltage, generating less heat and saving some electricity cost.
When playing with Frequency & Voltage (This is from S5 experience as I do not have an S7) then I usually aim for a HW error rate between .01% & .02% as I find this has no measurable effect on the Hash rate & because there are "some" errors I know that I am not wasting power.
Rich
When playing with Frequency & Voltage (This is from S5 experience as I do not have an S7) then I usually aim for a HW error rate between .01% & .02% as I find this has no measurable effect on the Hash rate & because there are "some" errors I know that I am not wasting power.
Rich
September 29, 2015, 05:20:21 AM
Can somebody tell me what the effect is of a high HW? If you have two different miners at the same speed, they hash about the same but one has a much higher HW. And on the pool there is also not much effect.
September 28, 2015, 08:13:48 PM
so the 1 unit I have thats been a pain in the ass so far has dropped to 3.9TH over the weekend (with no dead asic chips) and stock 600 settings.. going to try a hard reboot via PSU and see what happens
still tons of HW errors, im going to drop freq to 593 and see what happens. BITMAIN replied to me via email and asked what PSU im using, asked to try different pool (im using antpool, go figure) , and asked me to plug in each hash board individually and take screenshots of their hash rates..
only problem with that is it seems to take a couple days for the hashrate to crash on this unit, and I don't really want to waste a week of mining testing on each hash board individually.
hopefully the 593 freq will work out, i could live with 100GH loss instead of 900GH
That is exactly what they told me! One of mine is running at 4300th/s. Now I have to troubleshoot and see which board is causing the problem, then probably send it back to bitmain and after 3-4 weeks they will send me another board. Let me know your findings.
593 freq is running great so far! 4730GH/s avg at almost 7hrs with 0.026% HW ...ill take this all day long over the high HW and lower speed at stock settings, not sure how ours passed QC checks
try the lower freq and see what you get
September 28, 2015, 07:47:01 PM
so the 1 unit I have thats been a pain in the ass so far has dropped to 3.9TH over the weekend (with no dead asic chips) and stock 600 settings.. going to try a hard reboot via PSU and see what happens
still tons of HW errors, im going to drop freq to 593 and see what happens. BITMAIN replied to me via email and asked what PSU im using, asked to try different pool (im using antpool, go figure) , and asked me to plug in each hash board individually and take screenshots of their hash rates..
only problem with that is it seems to take a couple days for the hashrate to crash on this unit, and I don't really want to waste a week of mining testing on each hash board individually.
hopefully the 593 freq will work out, i could live with 100GH loss instead of 900GH
That is exactly what they told me! One of mine is running at 4300th/s. Now I have to troubleshoot and see which board is causing the problem, then probably send it back to bitmain and after 3-4 weeks they will send me another board. Let me know your findings.
September 28, 2015, 07:43:50 PM
Mine looks like dogie's with the pci-e outlet on the side of the controller.
It's been running for 4 days now and consistently between 4.810 - 4.830 at 600.
It's been running for 4 days now and consistently between 4.810 - 4.830 at 600.
September 28, 2015, 07:40:06 PM
Click to enlarge
Your picture is from the older version of S7, manufactured from S5+ leftover components. Controller and housing. Now we have a new and "better" S7.
Except this housing is difference to that of the S5+? The ones I saw certainly weren't extruded aluminium.
New S5+ housing is from aluminium. Exactly the same shape as your old S7 miner.
S5+ pictures
New S7 housing from philipma1957 picture
Click to enlarge
damn it looks like mine are the same as Dogie's review unit (the older style I guess?)
thats kind of a rip, I ordered day 1/2 .. do we know what "better" means on the newer S7?
the main tell is the controllers power connector on the older S7 is pointing out the side just like the S5+.. the newer S7 looks like its pointing up
I see the difference on controller on PCI-e going up.
Is there more details as far as the housing? I guess I would love to see side by sides if anyone happens to have both.
Looks pretty similar, just larger bulges and slots to link them together? I'm not sure we should really care, although my S7 seems to be clocking higher (675) than others are reporting?
September 28, 2015, 07:26:44 PM
I see the difference on controller on PCI-e going up. Does the "new" one have 9 slots for blades still I could not tell from picture looks like still nine i'm estimating after looking at it again..
Is there more details as far as the housing? I guess I would love to see side by sides if anyone happens to have both.
Yeah I think its 9 slots also on the new one.. but if you look close the little notches in the ribbon cables all go the same way, the older S7 and also the S5+ have the notches going different ways on each side of the controller
September 28, 2015, 07:20:56 PM
Click to enlarge
Your picture is from the older version of S7, manufactured from S5+ leftover components. Controller and housing. Now we have a new and "better" S7.
Except this housing is difference to that of the S5+? The ones I saw certainly weren't extruded aluminium.
New S5+ housing is from aluminium. Exactly the same shape as your old S7 miner.
S5+ pictures
New S7 housing from philipma1957 picture
Click to enlarge
damn it looks like mine are the same as Dogie's review unit (the older style I guess?)
thats kind of a rip, I ordered day 1/2 .. do we know what "better" means on the newer S7?
the main tell is the controllers power connector on the older S7 is pointing out the side just like the S5+.. the newer S7 looks like its pointing up
I see the difference on controller on PCI-e going up.
Is there more details as far as the housing? I guess I would love to see side by sides if anyone happens to have both.
September 28, 2015, 06:54:22 PM
Click to enlarge
Your picture is from the older version of S7, manufactured from S5+ leftover components. Controller and housing. Now we have a new and "better" S7.
Except this housing is difference to that of the S5+? The ones I saw certainly weren't extruded aluminium.
New S5+ housing is from aluminium. Exactly the same shape as your old S7 miner.
S5+ pictures
New S7 housing from philipma1957 picture
Click to enlarge
damn it looks like mine are the same as Dogie's review unit (the older style I guess?)
thats kind of a rip, I ordered day 1/2 .. do we know what "better" means on the newer S7?
the main tell is the controllers power connector on the older S7 is pointing out the side just like the S5+.. the newer S7 looks like its pointing up
September 28, 2015, 06:30:30 PM
hmm interesting.. im running both of them on a 2880w psu with the jabberwock breakout board
after 4 hours at 593freq im at 4720GH and 0.026% HW
after 4 hours at 593freq im at 4720GH and 0.026% HW
Could well be PSU variant, on a DPS-2000 I managed 650 perfectly stable with 2 PCI-Es and appeared to be the same on 675. 700 gave a low hash rate.
Jump to: